Machine learning interatomic potential with DFT accuracy for general grain boundaries in α-Fe DOI Creative Commons
Kazuma Ito, Tatsuya Yokoi,

Katsutoshi Hyodo

et al.

npj Computational Materials, Journal Year: 2024, Volume and Issue: 10(1)

Published: Nov. 13, 2024

Abstract To advance the development of high-strength polycrystalline metallic materials towards achieving carbon neutrality, it is essential to design in which atomic level control general grain boundaries (GGBs), govern material properties, achieved. However, owing complex and diverse structures GGBs, there have been no reports on interatomic potentials capable reproducing them. This accuracy for conducting molecular dynamics analyses derive guidelines. In this study, we constructed a machine learning potential (MLIP) with density functional theory (DFT) model energy, structure, arbitrary (GBs), including α-Fe. Specifically, employed training dataset comprising generated based crystal space groups. The GGB was evaluated by directly comparing DFT calculations performed cells cut near GBs from nano-polycrystals, extrapolation grades local environment active methods entire nano-polycrystal. Furthermore, analyzed GB energy structure α-Fe polycrystals through large-scale analysis using MLIP. average calculated MLIP 1.57 J/m 2 , exhibiting good agreement experimental predictions. Our findings demonstrate methodology constructing an representing GGBs high accuracy, thereby paving way computational science materials.

Language: Английский

nc-MeB2/a-MPEA structured (HfMoNbZr)xTi1-xBy films for enhanced hardness, toughness and tribological performance DOI Creative Commons
Jiyang Xie,

Yiming Ruan,

Hao Du

et al.

Surface and Coatings Technology, Journal Year: 2025, Volume and Issue: unknown, P. 131727 - 131727

Published: Jan. 1, 2025

Language: Английский

Citations

1
Random Sampling Versus Active Learning Algorithms for Machine Learning Potentials of Quantum Liquid Water DOI
Nore Stolte, János Daru, Harald Forbert

et al.

Journal of Chemical Theory and Computation, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 14, 2025

Training accurate machine learning potentials requires electronic structure data comprehensively covering the configurational space of system interest. As construction this is computationally demanding, many schemes for identifying most important structures have been proposed. Here, we compare performance high-dimensional neural network (HDNNPs) quantum liquid water at ambient conditions trained to sets constructed using random sampling as well various flavors active based on query by committee. Contrary common understanding learning, find that a given set size, leads smaller test errors not included in training process. In our analysis, show can be related small energy offsets caused bias added which overcome instead correlations an error measure invariant such shifts. Still, all HDNNPs yield very similar and structural properties water, demonstrates robustness procedure with respect algorithm even when few 200 structures. However, preliminary potentials, reasonable initial avoid unnecessary extension covered configuration less relevant regions.

Language: Английский

Citations

1
Machine-learning-potential-driven prediction of high-entropy ceramics with ultra-high melting points DOI Creative Commons
Hong Meng, Yiwen Liu, Hulei Yu

et al.

Cell Reports Physical Science, Journal Year: 2025, Volume and Issue: unknown, P. 102449 - 102449

Published: Feb. 1, 2025

Language: Английский

Citations

1
Predicting Mechanical and Thermal Properties of High‐Entropy Ceramics via Transferable Machine‐Learning‐Potential‐Based Molecular Dynamics DOI Open Access
Yiwen Liu, Hong Meng, Zijie Zhu

et al.

Advanced Functional Materials, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 15, 2024

Abstract The mechanical and thermal performance of high‐entropy ceramics are critical to their use in extreme conditions. However, the vast composition space significantly hinders development with desired properties. Herein, taking carbides (HECs) as model, efficiency effectiveness predicting properties via transferable machine‐learning‐potential‐based molecular dynamics (MD) have been demonstrated. Specifically, a neuroevolution potential (NEP) broad compositional applicability for HECs ten transition metal elements from group IIIB‐VIB is efficiently constructed small dataset comprising unary binary an equal amount ergodic chemical compositions. Based on this well‐established NEP, MD predictions different shown good agreement results first‐principles calculations experimental measurements, validating accuracy, transferability, reliability using simulations investigating HECs. This work provides strategy accelerate search desirable

Language: Английский

Citations

7
Accurate prediction of structural and mechanical properties on amorphous materials enabled through machine-learning potentials: A case study of silicon nitride DOI Creative Commons
Ganesh Kumar Nayak,

Prashanth Srinivasan,

Juraj Todt

et al.

Computational Materials Science, Journal Year: 2025, Volume and Issue: 249, P. 113629 - 113629

Published: Jan. 6, 2025

Language: Английский

Citations

0
Setting material benchmarks at large-strain limits via ultimate strengths DOI
Xinxin Gao, Kan Zhang, Qiang Zhu

et al.

Acta Materialia, Journal Year: 2025, Volume and Issue: 286, P. 120724 - 120724

Published: Jan. 9, 2025

Language: Английский

Citations

0
Combined X-ray microdiffraction and micromechanical testing for direct measurement of thin film elastic constants DOI Creative Commons
Rebecca Janknecht, Rainer Hahn, Nikola Koutná

et al.

Materials & Design, Journal Year: 2025, Volume and Issue: unknown, P. 113720 - 113720

Published: Feb. 1, 2025

Language: Английский

Citations

0
Transferable machine learning model for multi-target nanoscale simulations in hydrogen-carbon system from crystal to amorphous DOI Creative Commons
Weiqi Chen,

Zhiyue Xu,

Kang Wang

et al.

npj Computational Materials, Journal Year: 2025, Volume and Issue: 11(1)

Published: May 3, 2025

Language: Английский

Citations

0
Machine learning interatomic potential with DFT accuracy for general grain boundaries in α-Fe DOI Creative Commons
Kazuma Ito, Tatsuya Yokoi,

Katsutoshi Hyodo

et al.

npj Computational Materials, Journal Year: 2024, Volume and Issue: 10(1)

Published: Nov. 13, 2024

Abstract To advance the development of high-strength polycrystalline metallic materials towards achieving carbon neutrality, it is essential to design in which atomic level control general grain boundaries (GGBs), govern material properties, achieved. However, owing complex and diverse structures GGBs, there have been no reports on interatomic potentials capable reproducing them. This accuracy for conducting molecular dynamics analyses derive guidelines. In this study, we constructed a machine learning potential (MLIP) with density functional theory (DFT) model energy, structure, arbitrary (GBs), including α-Fe. Specifically, employed training dataset comprising generated based crystal space groups. The GGB was evaluated by directly comparing DFT calculations performed cells cut near GBs from nano-polycrystals, extrapolation grades local environment active methods entire nano-polycrystal. Furthermore, analyzed GB energy structure α-Fe polycrystals through large-scale analysis using MLIP. average calculated MLIP 1.57 J/m 2 , exhibiting good agreement experimental predictions. Our findings demonstrate methodology constructing an representing GGBs high accuracy, thereby paving way computational science materials.

Language: Английский

Citations

3